S&P Global announced an expansion of Xpressfeed last week so that its London-based market intelligence division could launch Textual Data Analytics (TDA). TDA, SPGI’s press literature reports, allows listed company investors to ‘incorporate more qualitative measures of company performance into their investment strategy by quantifying sentiment and behaviour during company calls.’
Active portfolio managers assess market value better than index funds about half the time, and BVWire—UK questions whether natural language processing of the ‘sentiment’ of company performance call will improve active manager performance. For the business valuation profession, this analysis could be another data point supporting the analyst’s judgment.
Many of the large business valuation firms in the UK rely on CapIQ, so they may have access to TDA as part of annual subscriptions for their analysts. ‘Our textual data analytics package of 40 behavioural and sentiment scores will further enrich our existing transcripts coverage, while offering clients a new alternative data set via our Xpressfeed delivery channel,’ says Warren Breakstone, managing director and chief product officer of data management solutions at S&P Global Market Intelligence (SPGMI).
TDA says that companies that exhibited good sentiment in their earnings and M&A calls outperformed the equities markets by over 2% a year between 2010 and 2017 (every model, when backtested, produces superior results—how else do you create an AI algorithm?).
This doesn’t mean analysts and valuers should change their guideline comparables formulas to reflect the ability of certain executives to project positive energy on earnings calls. This assessment is what valuers do every day, generally with private companies that do not hold such meetings. The key is understanding the business principles and economics, not the natural language elements, when using the income or market approaches to business valuation.
Please let us know
if you have any comments about this article or enhancements you would like to see.